BIOSYNTHETIC RADIOLABELING OF CELLULAR AND SECRETED PROTEINS OF MONONUCLEAR PHAGOCYTES

BIOSYNTHETIC RADIOLABELING OF CELLULAR AND SECRETED PROTEINS OF MONONUCLEAR PHAGOCYTES

Zena Werb Jennie R. Chin

I. GENERAL INTRODUCTION

A complete description of the phenotype of mononuclear phagocytes involves defining all the biochemical and function- al properties of the cells. One approach is to examine the pattern of total transcribable mRNA present in a macrophage at any given point in its history. This can be achieved either by translating isolated mRNA in cell-free systems, a process unsuitable for examination of many samples, or by studying the translation of mRNA into proteins using the machinery of a live cell. If radiolabeled amino acids are present during translation in live cells, the resulting bio-

synthesized proteins can be analyzed by polyacrylamide gel electrophoresis to give detailed "fingerprints" of specific macrophage phenotypes (1,2). These procedures offer high resolution and specificity, require small numbers of cells

(as few as 1 x 10

), and are applicable to a wide variety of mononuclear phagocytes from man, mouse, rat, rabbit, and

METHODS FOR STUDYING Copyright © 1981 by Academic Press, Inc.

MONONUCLEAR PHAGOCYTES 861 All rights of reproduction in any form reserved.

ISBN 0-12-044220-5

guinea pig. In conjuction with other methods, such as speci- fic immunoprecipitation of labeled proteins, it is possible to examine in detail changes in the properties of macrophages that may not be seen with a single assay such as receptor binding or quantitation of a secreted proteolytic enzyme. In addition, these methods can be used to detect contamination by other cell populations by searching for biosynthesized proteins specific for those cells.

II. [^JDS]-METHIONINE LABELING AND SAMPLE PREPARATION OF SECRETED AND CELLULAR PROTEINS OF MACROPHAGES

A. Introduction

Resident or elicited macrophages are harvested from the peritoneal cavity or lungs by lavage with phosphate-buffered saline containing 100 U/ml heparin (see Section II of this volume). When rabbit alveolar macrophages are to be studied, it is essential to isolate and plate them as quickly and gently as possible to prevent clumping. Mouse cells are less apt to undergo this aggregation.

B. Reagents

All tissue culture media are obtained from Grand Island Biological Co., Grand Island, New York, and stored at 4°C except where noted. Culture plasticware is available from standard suppliers (Flow Laboratories, M. A. Bioproducts, GIBCO, Costar).

(1) Heat-inactivated fetal calf serum. Thaw to room temperature and heat in a 56°C water bath for 30 to 60 min.

(2) Lactalbumin hydrolysate (extra-soluble, tissue cul- ture grade) (LH). Sterilize a 10% stock solution in water by filtration through a 0.22-ym filter. Store at -20°C.

(3) Culture medium. Dulbecco's modified Eagle's medium (DME) is supplemented with heat-inactivated fetal calf serum to a final concentration of 10% (DME-10%) or with LH (DME-LH) to 0.2%. Add penicillin-streptomycin solution to a final con- centration of 50 U/ml penicillin and 50 yg/ml streptomycin.

RPMI-1640 or minimal essential medium can be substituted for DME.

Methionine-free medium can be purchased as a custom order from GIBCO; or minimal essential medium and RPMI can be mixed from "select-amine" kits available from GIBCO.

(4) [^JbS]-methionine. Amersham, >1000 Ci/mmole. Store at -80°C until needed. Labeling medium is made up fresh at 25 yCi/ml in methionine-free medium.

(5) Micrococcus lysodeikticus. Sigma. Wash dried cells once with saline and make up at 2 mg/ml with saline. The suspension is stored at 4°C and should be made up fresh every 2 weeks; prolonged storage may result in hydrolysis of the cell suspension.

(6) Trichloroacetic acid (45% w/v). This is used for precipitation of labeled proteins.

(7) 2X Sample buffer. Section III. B.

(8) Sample tubes. Screwcapped cryotubes, 2- and 5-ml sizes, manufactured by Nunc or Cooke, are available from GIBCO. Microfuge tubes (1.5 ml polypropylene) are available from Beckman and other suppliers.

C. Procedures

(1). Plate the macrophages in DME-10% at 5 x 1 0⁵ to 1 0⁶ cells per well in 12-well Linbro plates (2.4-cm diameter) or 24-well Costar plates (16-mm diameter). Allow cells to ad- here 2-24 hr at 37°C in 5% C 02 in humidified air. If the cells are to be treated with drugs, proceed to the next step;

otherwise, go to step 4 for direct labeling.

(2). Wash the cells at least three times with DME-LH to remove serum and nonadherent cells. Phosphate-buffered saline or Hanks¹ balanced salt solution can also be used for prelimi- nary washes.

(3). Pretreat the macrophages with 0.5-1.0 ml of DME-LH or DME-10% containing the drug for the desired time period.

Stock solutions of drugs are usually prepared 100X concentra- ted in methionine-free DME; ethanol content, if any, should be 0.05% or less to maintain cell viability.

(4). Wash the macrophages three times with methionine- free DME and label with 0.5-1.0 ml methionine-free DME con- taining 25 yCi [³⁵S]-methionine per ml; usually any druq used during pretreatment is also added. Macrophages are generally labeled for 2-4 hr, although periods as short as 10 min are feasible for special purposes. Up to 300 yCi [³^S]-methionine per ml may be needed for very short labeling times.

(5). At the end of the labeling period, collect the conditioned medium in 1.5-ml microfuge tubes with Pasteur pipettes. Wash the cells 3-4 times with saline and lyse (in the plate) with 0.2 to 0.5 ml of IX sample buffer or 0.1%

sodium dodecyl sulfate (SDS) followed by an equal volume of 2X sample buffer. Freeze plates at -80°C in ziplock plastic

METHODS FOR STUDYING MONONUCLEAR PHAGOCYTES bags. Just before electrophoresis, check for complete lysis by microscopy and transfer the lysates to microfuge tubes or cryotubes for boiling (step 11).

(6). Spin the collected medium 2.5 min in a Beckman Microfuge B (approx. 8730 g) to remove any loose cells and debris. This step is of utmost importance for analysis of macrophage secretion products; otherwise, minor cellular con- taminants will appear in the fluorographs along with proteins from the medium. Presence of actin (42,000 daltons) in con- ditioned medium is a marker for contamination by cell debris.

(7). With a Pasteur pipette carefully transfer all but the bottom 10 yl of medium to another microfuge tube; no pellet will be visible.

(8). To each sample, add 150 yg (75 yl) M. lysodeikticus cell suspension (kept stirring on a magnetic stirrer) as a

carrier. This carrier was chosen because it does not enter the gel during electrophoresis, as would protein carriers such as bovine serum albumin, and, consequently, there is less chance of interference with the electrophoretic run.

Precipitate proteins with ice-cold trichloroacetic acid (150 yl of 45% w/v) to a final concentration of 5-7%. Cap and shake tubes and let stand in an ice bath for at least 20 min.

(9). Centrifuge, wash, and vortex the M. lysodeikticus- containing pellet 1-2 times with 1 ml 5% trichloroacetic acid

or imlacetone. Remove as much supernatant as possible with a Pasteur pipette and discard the supernatant as radioactive waste.

(10). Vortex the pellet in 100 yl of IX sample buffer, if necessary adjust to alkaline pH with 1-5 yl 1 N NaOH (yellow, acidic sample becomes blue when basic). Caution: Excessively high concentrations of NaOH (pH greater than 10) when boiled will result in hydrolysis of the protein sample, M. Lyso-

the day of electrophoresis.

(11). Place tightly capped sample tubes into a boiling water bath for 3 min. Microfuge tubes have a tendency to pop open during boiling unless a weight is placed on their lids.

An alternative is to use screwtop cryotubes.

(12). Determine incorporation by putting 5-yl aliquots

into counting vials containing scintillant (work in the fume

hood), rinsing out the pipette tip with the scintillant. This

precaution is necessary because the glycerol and SDS-contain-

ing sample tends to adhere to the walls of the tip, resulting

in incorrect counts.

III. SDS-POLYACRYLAMIDE GRADIENT GEL ELECTROPHORESIS AND FLUOROGRAPHY OF LABELED MACROPHAGE PROTEINS

A. Introduction

Samples are electrophoresed on 7-18% polyacrylamide - 0.1%

SDS gradient slab gels, using the Tris-glycine-SDS buffer sys- tem of Laemmli (3). The resolving power of gradient gels is far superior to that of uniform percentage gels, which tend to yield fuzzier bands, especially with proteins in the condition- ed medium. For some types of analysis it may be helpful to add 2 M urea to gels and sample buffer.

Low plating efficiencies and short labeling periods will obviously result in lower incorporation. Two fluorographic methods are presented to increase the efficiency of detection of bands up to tenfold for *^^S and ^ C a nd to allow visualiza- tion of ^H-labeled proteins.

B. Reagents

All electrophoresis stock solutions are made from electro- phoretic grade reagents from Bio-Rad. Many designs of elec- trophoresis apparatus can be used. A suitable design is avail- able from Bio-Rad.

(1) 30% Acrylamide-W ,iV'-methylenebisacrylamide.

Dissolve 29.2 gm acrylamide and 0.8 gm bisacrylamide in triple distilled water, adjust the volume to 100 ml, and filter through a No. 1 Whatman filter paper. Store at 4°C in a brown bottle. Avoid contact with skin; acrylamide in solution is a neurotoxin.

(2) 4X Lower gel buffer. 1.5 M Tris-HCl, pH 8.8, with 0.4% SDS. Store at 4°C.

(3) 4X Upper gel buffer. 0.5% M Tris-HCl, pH 6.8, with 0.4% SDS. Store at 4°C.

(4) 10% Ammonium persulfate. Make up fresh in water.

(5) Ν,Ν,Ν',iV'-Tetramethylethylenediamine (TEMED). Store at 4°C. Prolonged gelling time may be due to weak or old

TEMED.

(6) 10X Electrode buffer. 0.25 Ai Tris, 1.9 M glycine, and 1% SDS. Store at room temperature. Dilute with water before use.

(7) 2X Sample buffer. 1% ß-mercaptoethanol, 0.1% bromo- phenol blue, 0.0625 M Tris-HCl, pH 6.8, 50% glycerol, and 2%

SDS. Store at room temperature. For IX sample buffer, dilute with water.

(8) ^{1 4}OMethylated protein markers. Amersham. Contains myosin, phosphorylase Bf bovine serum albumin, ovalbumin, car- bonic anhydrase, and lysozyme (0.833 yCi/ml each).

(9) EN-HANCE. New England Nuclear. Store at 4°C.

(10) Kodak X-Omat AR X-ray film.

C. Procedures

(1). Gradient slab gels are prepared a day in advance of electrophoresis. The volumes needed depend on the design of the electrophoresis apparatus; best results are obtained with gels that are 0.75-mm thick and 10-20 cm long. For each slab gel containing 0.375 M Tris-HCl, pH 8.8, and 0.1% SDS, prepare one volume each of 7% and 18% acrylamide gel solution accord- ing to the following recipe: Mix gently in a beaker on ice:

0.25 vol lower gel buffer, 0.75 vol acrylamide - water (dilute stock acrylamide with appropriate volume of water to give a final concentration of 7% or 18%), 0.00145 vol 10% ammonium persulfate, and 0.0005 vol TEMED. A few grains of bromophenol blue are added to the 7% solution to aid in visualization of gradient formation and water overlaying.

(2). Pour the gradient gel with a linear gradient maker at a rate of 1 ml/min. Gently overlay the surface with water from a 22-gauge, beveled 2-in. needle and syringe. Let poly- merize undisturbed for 20-30 min. Prolonged polymerization

time is indicative of old TEMED and fresher catalyst should be used. Wash the surface twice with fresh water. The gel may be left at room temperature overnight. For longer storage

(2-3 days) use IX lower gel buffer to overlay the gel.

(3). Pour off the water and gently blot the surface with bibulous paper before adding the stacking gel solution. For a 3% stacker gel containing 0.125 M Tris-HCl, pH 6.8, and 0.1%

SDS, mix gently in a beaker: 0.25 vol upper gel buffer, 0.1 vol acrylamide, 0.64 vol water, 0.01 vol 10% ammonium persulfate, and 0.0005 vol TEMED. Quickly pipette the solution onto the gradient gel and insert the sample-well comb, without forming any air pockets. Sample wells 1.8-2.8 cm long and 0.45-0.9 cm wide (10- or 20-well combs) are usually used. Let polymerize. Wash the wells twice with IX electrode buffer.

(4). Apply samples to the wells with either equivalent counts or proportionate volumes. The former will show changes in incorporation by particular bands, whereas the latter will show overall changes in the whole secretory pattern. Reserve one well for ^-⁴C-labeled protein markers (diluted 1 to 40 with IX sample buffer and boiled 3 min.).

If different volumes are applied, a more "even" run is achieved by equalizing the volumes in each well with half- strength sample buffer before overlaying with electrode buffer.

(5). Electrophorese at 20 mA per gel. A typical run time for 10 x 14 x 0.75 mm gels is 3 hr.

(6). Fix the gels in 50% trichloroacetic acid for 1 hr or overnight. A staining step is usually eliminated because there is not enough protein to be detected visually. Any staining would be extracted by preparation for fluorography.

(7). One of two procedures can be used to fluorograph gels. Bonner and Laskey's method (4) involves replacement of water in the gel with dimethyl sulfoxide (DMSO), impregnation with the fluor, 2,5-diphenyloxazole (PPO), followed by preci- pitation with water. To summarize briefly:

a. Working in a fume hood and wearing gloves, immerse the fixed gel in 20 vol of DMSO for 30 min. Repeat with fresh DMSO.

b. Soak the gel for 3 hr in 22.2% (w/v) PPO in DMSO.

Discard this solution after one use.

c. Precipitate the PPO by submerging the gel in water for 1 hr.

d. Dry the gel under vacuum on a standard slab gel drying apparatus.

The second procedure eliminates steps 1 and 2 and consists of soaking the gel for 1 hr in a commercial fluorographic solu- tion, EN^HANCE, followed by precipitation with water for 1 hr.

(8). Preflash Kodak X-Omat AR X-ray film before exposure to the gel. According to Laskey and Mills (5), the absorbance of the fluorographic image is not linear with the amount of radioactivity or with exposure time. Preexposure of the film to a flash of light from a flash unit corrects for this non- linearity and increases the efficiency of detection. Place the film against the gel and seal them inside the film envelope with photographic tape. If necessary, wrap the envelope with lead foil to prevent exposure from any stray radioactive sources. Secure between two boards with clamps and expose at -80°C.

IV. CALCULATION OF DATA

A. Introduction

Fluorographs can be scanned in a gel densitometer (Joyce-

Loebl, Canalco, etc.) to determine molecular weights and rela-

tive incorporation by individual bands.

B. Procedures

(1). With a sharp pencil draw a line approximately 1-2 mm above the origin and below the front of the lane to be scanned.

(Placing millimeter graph paper behind the film is very help- ful.) This step produces a sharp peak and shows the exact be- ginning and end of the scan.

(2). Peak areas can be readily calculated by those den- sitometers equipped with integrators. An alternative method is to make several xerographic copies of the scan and to re- solve and extend each peak down to the baseline. Each one is then cut out exactly and weighed on a Mettler balance. Knowing the individual peak weights, the total weight of all the peaks, and the number of counts in the applied sample, it is possible to calculate the percentage of the total incorporation and the absolute number of counts in a single band. For example:

Weight of all peaks = 1500 mg

Peak X = 100 mg = 6.7% of total weight Counts in applied sample = 15,000 cpm

Therefore, counts in Peak X = 6.7% (15,000) = 1000 cpm (3). Individual bands, once localized, can be cut out of the original gel, solubilized, and counted in a scintillation spectrometer for additional quantitation.

(4). Molecular weights can be determined by comparison of the R

values of each band with those of the standards (6).

V. CRITICAL COMMENTS

A. Expected Incorporation

Incorporation of [^^S]-methionine into cell-associated and secreted proteins varies with cell number and labeling time, as well as time in culture (see Section V. B. 2). Protein patterns from 10^ macrophages can be detected in as little as 15 min of labeling; however, it should be noted that the ex- posure time will be longer than that for cells labeled 2-4 hr.

If fewer cells are available, e.g., 10^, the incorporation time can be increased to 4 hr and fluorographs can be readily examined after only 10 days of exposure at -80°C.

The total incorporation into the secreted proteins of

macrophages varies from 1.5 to 20% of the total incorporated

label. Typical labeling of 5 x 10

thioglycollate-elicited

macrophages in a 2-hr labeling period under the conditions

described here is approximately 2.3 x 10

cpm in the secreted

proteins and about 1 x 10^ cpm in the cell-associated proteins.

For similar numbers of resident macrophages, approximately 1 x 1θ4 cpm are seen in the secreted proteins and about 1.2 x 10

cpm in the cell-associated proteins.

B. Reproducibility of Incorporation

The reproducibility of incorporation depends on the age and specific activity of [^S] -methionine and the time in cul- ture of the macrophages.

(1). Because the half-life of

S is only 87.5 days, the concentration of the isotope is recalculated with each use to maintain a constant concentration from experiment to experi- ment. It should also be noted that storage of radioisotope for 3 weeks at 20°C results in 50% loss of radiochemical puri- ty. By 6 weeks, over 75% of [

S]-methionine has decomposed to [35s]-methionine sulfoxide. Therefore, Amersham recommends storage at or below -80°C.

[

S]-methionine is commercially available with specific activity of 1000 Ci/mmol or 500 Ci/mmol. To get maximum in- corporation, label with the former.

(2). Preliminary experiments have shown that incorporation by macrophages varies with time in culture (unpublished results) For example, NaI0

-elicited mouse peritoneal macrophages were cultured in DME-10% for either 2 hr or 2 days (medium was changed once) before labeling. Cells in the 2-day culture in- corporated approximately 30% less label in their secreted pro- teins than those in the 2-hr culture. With macrophages elicit- ed by other means, incorporation was higher in the older cul- tures.

A comparison of fluorographs showed loss or change in a number of bands. Although this phenomenon needs to be investi- gated further, it is preferable to label macrophages that have been explanted into culture for less than 72 hr and to be con- sistent with culture time.

C. Sensitivity

Metabolic labeling with [^^S]-methionine is a very sensi- tive analytic technique for the study of both secreted and cellular proteins from macrophages. The high energy of 35s combined with its short half-life and its high specific acti- vity relative to -^^C- and

H-labeled amino acids allows greater incorporation. In addition, fluorography increases the effi- ciency of detection tenfold over conventional autoradiography.

Bonner and Laskey (4) were able to detect 0.06 nCi/band after

24-hr exposure at -70°C. Exposure time will, of course, depend

on the total number of bands present, as well as the number of

870

counts applied. For example, 2 5,000 cpm distributed among 30 bands of secreted proteins were visible after 4 days of exposure at -70°C.

D. Differences in Synthesis and Secretion Patterns in Mononuclear Phagocytes from Different Sources

The procedures outlined here are powerful tools in tracing the history of mononuclear phagocytes. We have found that the gel patterns of both cell-associated and secreted proteins of macrophages from various inbred mouse strains, including the recombinant resistant strains, exhibited significant differences that cannot be attributed to differences in the H-2 haplotype of the mice. The most striking distinctions observed were be- tween resident and inflammatory macrophages, and several sub- classes of inflammatory macrophages (e.g., thioglycollate-, endotoxin-, and pyran copolymer-elicited) can be differentiated by using [35g]-

thionine, [^H]-mannose as labels. There are similarities between species: rat, rabbit, human, and mouse macrophages share some major secreted proteins but also differ

significantly in others. Similarly, proliferating tumor-de- rived mouse macrophages share some major proteins with macro- phages from normal mice. However, in this case, major differ- ences are possibly due to the biosynthesis and secretion of viral proteins by these cells. These labeling procedures can also be used to examine modulation of macrophage function.

For example, new proteins are biosynthesized in response to treatment of mouse macrophages with glucocorticoids, prosta- glandins, and antibody-coated erythrocytes.

Another important use of this technique is to help evaluate the contamination of macrophages or other cell populations by other cells. For example, the presence of radiolabeled plasma proteins such as albumin and fibrinogen in the gel patterns of secreted proteins from Kupffer cell populations is indicative of hepatocyte contamination. Similarly, fibroblast-specific proteins can be used to evaluate fibroblast contamination in inflammatory macrophage monolayers, and similar evaluations can be made for thymocytes and B cells.

E. Modifications

The procedures outlined in Sections II and III involve the use of [^^S]-methionine as the biosynthetic label of choice.

However, other isotopes, such as [^H]-mannose, [^Hl-glucosamine, other amino acids, and even ^^P, can be used in this way (7-11).

Indeed, the most dramatic differences between resident and

inflammatory macrophages appear to be specific to the secreted glycoproteins. Carbohydrate labels are particularly useful in this regard.

Specific labeled proteins can be immunoprecipitated direct- ly with immune serum or indirectly with specific immunoglobu- lins followed by protein A or a second antibody (12). These techniques have been applied to the investigation of macrophage o^-macroglobulin (13), fibronectin (13) , complement components

(9,10,14), and ß-glucuronidase (11).

Additional details about the biosynthesis and processing of macrophage proteins can be obtained by peptide mapping of specific proteins as described by Cleveland et al. (15) and Johnson et al. (16). These techniques have been applied to studies of lysosomal enzymes and complement proteins (9,11).

Acknowledgment

This work was supported by the U. S. Department of Energy.

REFERENCES

1. Z. Werb and J. Chin. The secretion phenotypes of resident and inflammatory mouse macrophages. Manuscript in prepara- tion.

2. Z. Werb. The effects of toxic chemicals on the immune system: The interaction of macrophages with glucocorti- coids as a model system. In "Proceedings of the 10th Conference on Environmental Toxicology," Dayton, Ohio, 13-15 November 1979, pp. 154-171, 1980.

3. U. K. Laemmli. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685, 1970.

4. W. M. Bonner and R. A. Laskey. A film detection method for tritium-labeled proteins and nucleic acids in poly- acrylamide gels. Eur. J. Biochem. 46:83-88, 1974.

5. R. A. Laskey and A. D. Mills. Quantitative film detection of ³H and ^{1 4}C in polyacrylamide gels by fluorography.

Eur. J. Biochem. 56:335-341, 1975.

6. K. Weber and M. Osborn. The reliability of molecular weight determinations by dodecyl sulfate-polyacrylamide gel electrophoresis. J. Biol. Chem. 244:4406-4412, 1969.

7. D. K. Struck, P. B. Siuta, M. D. Lane, and W. J. Lennarz.

Effect of tunicamycin on the secretion of serum proteins by primary cultures of rat and chick hepatocytes. J. Biol.

Chem. 253:5332-5337, 1978.

8. A. Tartakoff, P. Vassalli, and M. Detraz. Comparative studies of intracellular transport of secretory proteins.

J. Cell Biol. 79:694-701, 1978.

9. J. E. Pennington, W. J. Matthews. Jr., J. T. Marino, Jr., and H. R. Colten. Cyclophosphamide and cortisone acetate inhibit complement biosynthesis by guinea pig broncho- alveolar macrophages. J. Immunol. 123:1318-1321, 1979.

10. C. Bentley, D. Bitter-Suermann, U. Hadding, and V. Brade.

In vitro synthesis of factor B of the alternative path- way of complement activation by mouse peritoneal macro- phages. Eur. J. Immunol. 6:393-398, 1976.

11. M. D. Skudlarek and R. T. Swank. Biosynthesis of two lysosomal enzymes in macrophages. J. Biol. Chem. 254:

9939-9942, 1979.

12. S. W. Kessler. Cell membrane antigen isolation with the staphylocococcal protein A-antibody adsorbent. J. Immu- nol. 117:1482-1490, 1976.

13. T. Hovi, D. Mosher, and A. Vaheri. Cultured human raono- cytes synthesize and secrete a2-macroglobulin. J. Exp.

Med. 145:1580-1589, 1977.

14. M. H. Roos, J. P. Atkinson, and D. C. Shreffler. Mole- cular characterization of the Ss and Sip (C4) proteins of the mouse H-2 complex: Subunit composition, chain size polymorphism, and an intracellular (Pro-Ss) pre- cursor. J. Immunol. 121:1106-1115, 1978.

15. D. W. Cleveland, S. G. Fischer, M. W. Kirschner, and U. K. Laemmli. Peptide mapping by limited proteolysis in sodium dodecyl sulfate and analysis by gel electro- phoresis. J. Biol. Chem. 252:1102-1106, 1977.

16. E. F. Johnson, M. C. Zounes, and U. Muller-Eberhard.

Characterization of three forms of rabbit microsomal cytochrome P-450 by peptide mapping utilizing limited proteolysis in sodium dodecyl sulfate and analysis by gel electrophoresis. Arch. Biochem. Biophys. 192:282- 289, 1979.